The spy shot the cop with the revolver.

This sentence, a favorite of linguists, appears to be simple enough. It's grammatically correct, has a subject and a predicate and can even be easily understood by young children. Or can it? Who had the revolver: the spy or the cop?

Like optical illusions, language can play tricks on the brain, explained New York University psychologist Gary Marcus at an April 6 lecture at the American Museum of Natural History in New York. These simple syntactic ambiguities, he argued, throw a wrench into prevalent theories that the human brain is well evolved—or even optimally evolved—for language.

As evidence for our species's general optimality, many people point to the complex human visual system, which has yet to be matched by technological developments. So if our brain's handling of vision is so well tuned, shouldn't our language centers be, as well?

No way, according to Marcus. Visual abilities have been developing in animal predecessors for hundreds of millions of years. Language, on the other hand, has had only a few hundred thousand years to eke out a place in our primate brain, he noted. What our species has come up with is a "kluge," Marcus said, a term he borrows from engineering that means a solution that is "clumsy and inelegant, but it gets the job done." (His 2008 book, Kluge: The Haphazard Construction of the Human Mind, explores the language subject and other human mental inefficiencies.)

Unlike a computer, which can work with C++ and other languages that lack the ambiguities of say, English or Chinese, the human brain's capacity to understand language is tied up in how our memory works—and that is deeply rooted in evolutionary pressures, stuck in what he called "evolutionary inertia."

The human brain has a very context-driven memory (unlike computers which have location-based memory), which is why people often have trouble remembering where their car is in a parking lot or parking garage they use frequently, Marcus explained. These mental contextual cues, however, are not ideal for an organized, rational development—or use—of language.

In fact, one sociologist tried to make just such a logical language. The language, called Loglan, was described by its creator James Cook Brown in a 1960 article [pdf] in Scientific American. Unfortunately for Brown, who used mathematic and logic models to build the language, the experiment failed in that people actually had a very difficult time learning the artificial tongue.

But this failure boosts Marcus's idea that natural human languages are not a product of the well-orchestrated rationality and neat syntactic trees like Noam Chomsky and other linguists have described.

Rather they are messy approximations that manage to get people's point across. Most of the the time.

Read a conversation with Marcus about his book and research.

Image courtesy of iStockphoto/megatronservizi